Consideration to apply the Mahalanobis-Taguchi method to small sample data

Masato Ohkubo, Yasushi Nagata

    Research output: Contribution to journalArticle

    2 Citations (Scopus)

    Abstract

    The Mahalanobis-Taguchi (MT) method is a standard method of multivariate analysis for detecting anomalies or recognizing patterns. A number of case studies using the MT method have been reported. However, good performance is only obtained when a sufficient number of samples can be ensured; if the number of samples is insufficient, this method has a large probability bias. In this paper, we first analyze the existing measures of methods, in which performing dimension reduction, such as using variable selection, is common, and show that there are some problems with testing for unknown data. Secondly, we propose two analytical procedures for small sample data in which the detection capability with respect to unknown data is taken into account. In these proposed procedures, when the number of data samples is small compared to the dimensions of the variables, the detection measure in the MT method is replaced by a measure derived through approximating correlation matrices based on probabilistic principal component analysis (PPCA) or by introducing ensemble learning. Finally, based on raw data analysis using the KDDCup99 dataset and simulation results, we consider how the proposed procedures should be applied when multicollinearity occurs and which of these two procedures should be applied according to the data pattern.

    Original languageEnglish
    Pages (from-to)30-38
    Number of pages9
    JournalJournal of Japan Industrial Management Association
    Volume66
    Issue number1
    Publication statusPublished - 2015

    Fingerprint

    Taguchi Method
    Taguchi methods
    Small Sample
    Principal component analysis
    Ensemble Learning
    Multicollinearity
    Unknown
    Probabilistic Analysis
    Multivariate Analysis
    Correlation Matrix
    Dimension Reduction
    Variable Selection
    Principal Component Analysis
    Testing
    Anomaly
    Data analysis
    Taguchi method
    Small sample
    Sufficient
    Simulation

    Keywords

    • Ensemble learning
    • MT method
    • Probabilistic principal component analysis
    • Taguchi method

    ASJC Scopus subject areas

    • Industrial and Manufacturing Engineering
    • Applied Mathematics
    • Management Science and Operations Research
    • Strategy and Management

    Cite this

    Consideration to apply the Mahalanobis-Taguchi method to small sample data. / Ohkubo, Masato; Nagata, Yasushi.

    In: Journal of Japan Industrial Management Association, Vol. 66, No. 1, 2015, p. 30-38.

    Research output: Contribution to journalArticle

    @article{8612d7f57fc4450681493827be28367f,
    title = "Consideration to apply the Mahalanobis-Taguchi method to small sample data",
    abstract = "The Mahalanobis-Taguchi (MT) method is a standard method of multivariate analysis for detecting anomalies or recognizing patterns. A number of case studies using the MT method have been reported. However, good performance is only obtained when a sufficient number of samples can be ensured; if the number of samples is insufficient, this method has a large probability bias. In this paper, we first analyze the existing measures of methods, in which performing dimension reduction, such as using variable selection, is common, and show that there are some problems with testing for unknown data. Secondly, we propose two analytical procedures for small sample data in which the detection capability with respect to unknown data is taken into account. In these proposed procedures, when the number of data samples is small compared to the dimensions of the variables, the detection measure in the MT method is replaced by a measure derived through approximating correlation matrices based on probabilistic principal component analysis (PPCA) or by introducing ensemble learning. Finally, based on raw data analysis using the KDDCup99 dataset and simulation results, we consider how the proposed procedures should be applied when multicollinearity occurs and which of these two procedures should be applied according to the data pattern.",
    keywords = "Ensemble learning, MT method, Probabilistic principal component analysis, Taguchi method",
    author = "Masato Ohkubo and Yasushi Nagata",
    year = "2015",
    language = "English",
    volume = "66",
    pages = "30--38",
    journal = "Journal of Japan Industrial Management Association",
    issn = "0386-4812",
    publisher = "Nihon Keikei Kogakkai",
    number = "1",

    }

    TY - JOUR

    T1 - Consideration to apply the Mahalanobis-Taguchi method to small sample data

    AU - Ohkubo, Masato

    AU - Nagata, Yasushi

    PY - 2015

    Y1 - 2015

    N2 - The Mahalanobis-Taguchi (MT) method is a standard method of multivariate analysis for detecting anomalies or recognizing patterns. A number of case studies using the MT method have been reported. However, good performance is only obtained when a sufficient number of samples can be ensured; if the number of samples is insufficient, this method has a large probability bias. In this paper, we first analyze the existing measures of methods, in which performing dimension reduction, such as using variable selection, is common, and show that there are some problems with testing for unknown data. Secondly, we propose two analytical procedures for small sample data in which the detection capability with respect to unknown data is taken into account. In these proposed procedures, when the number of data samples is small compared to the dimensions of the variables, the detection measure in the MT method is replaced by a measure derived through approximating correlation matrices based on probabilistic principal component analysis (PPCA) or by introducing ensemble learning. Finally, based on raw data analysis using the KDDCup99 dataset and simulation results, we consider how the proposed procedures should be applied when multicollinearity occurs and which of these two procedures should be applied according to the data pattern.

    AB - The Mahalanobis-Taguchi (MT) method is a standard method of multivariate analysis for detecting anomalies or recognizing patterns. A number of case studies using the MT method have been reported. However, good performance is only obtained when a sufficient number of samples can be ensured; if the number of samples is insufficient, this method has a large probability bias. In this paper, we first analyze the existing measures of methods, in which performing dimension reduction, such as using variable selection, is common, and show that there are some problems with testing for unknown data. Secondly, we propose two analytical procedures for small sample data in which the detection capability with respect to unknown data is taken into account. In these proposed procedures, when the number of data samples is small compared to the dimensions of the variables, the detection measure in the MT method is replaced by a measure derived through approximating correlation matrices based on probabilistic principal component analysis (PPCA) or by introducing ensemble learning. Finally, based on raw data analysis using the KDDCup99 dataset and simulation results, we consider how the proposed procedures should be applied when multicollinearity occurs and which of these two procedures should be applied according to the data pattern.

    KW - Ensemble learning

    KW - MT method

    KW - Probabilistic principal component analysis

    KW - Taguchi method

    UR - http://www.scopus.com/inward/record.url?scp=84931080651&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84931080651&partnerID=8YFLogxK

    M3 - Article

    AN - SCOPUS:84931080651

    VL - 66

    SP - 30

    EP - 38

    JO - Journal of Japan Industrial Management Association

    JF - Journal of Japan Industrial Management Association

    SN - 0386-4812

    IS - 1

    ER -